fix(tiff): decode UserComment EXIF tag (Bug #19)

[rpuneet]

Summary

• Decodes UserComment EXIF tag (0x9286) from raw bytes to readable text
• Supports ASCII, Unicode (UTF-16 LE/BE), JIS, and undefined charsets
• Handles edge cases: empty content, trailing nulls, malformed data, no prefix

Before/After

Before	After
415343494900000053686f7420...	"Shot during golden hour, 3-stop graduated ND filter used"
{"size": 264, "type": "binary"}	Decoded text string

Test plan

• Unit tests for all encoding types (ASCII, Unicode, JIS, undefined)
• UTF-16 BOM detection (LE/BE)
• Edge cases (empty, malformed, trailing nulls)
• Integration test updated
• All existing tests pass

Changes

File	Lines	Description
types.go	+1	Added TagUserComment constant
ifd.go	+160	decodeUserComment(), decodeUTF16(), isValidUTF8()
ifd_test.go	+208	24 test cases
api_integration_test.go	+1	Updated expectation

Closes #19

🤖 Generated with Claude Code

[rpuneet]

tech-lead-02 Review: LGTM ✅

Implementation matches the guidance I provided earlier. Well done!

Verified:

✅ Charset prefix detection (ASCII, Unicode, JIS, undefined)
✅ UTF-16 BOM detection (LE/BE)
✅ Surrogate pair handling for characters outside BMP
✅ Fallback for cameras without prefix
✅ 24 test cases covering edge cases

Clean implementation. Ready for merge once CI passes.

@tech-lead-01: Can you approve?

[codecov]

Codecov Report

❌ Patch coverage is 75.94937% with 19 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
internal/parser/tiff/ifd.go	75.94%	15 Missing and 4 partials ⚠️

📢 Thoughts on this report? Let us know!